• No results found

Implementation of Additive Manufacturing in Uprights for a Formula Student Car

N/A
N/A
Protected

Academic year: 2021

Share "Implementation of Additive Manufacturing in Uprights for a Formula Student Car"

Copied!
49
0
0

Loading.... (view fulltext now)

Full text

(1)

Implementation of Additive Manufacturing in Uprights for a Formula Student Car

SVEN-RUBEN BÖCKER KAJETAN CALCZYNSKI SIMON MALMSTRÖM

Bachelor Thesis Stockholm, Sweden 2016

(2)
(3)

Implementation of Additive Manufacturing in Uprights for a Formula Student Car

Sven-Ruben Böcker Kajetan Calczynski

Simon Malmström

Examensarbete MMKB 2016:49 MKNB 088 KTH Industriell teknik och management

Maskinkonstruktion SE-100 44 STOCKHOLM

(4)
(5)

Sammanfattning

Detta kandidatexamensarbete fokuserar på möjligheterna att implementera additiv tillverkning på en styrspindel, en av nyckelkomponenterna i en Formula Student-bil. Målet var att få en inblick i denna tillverkningsteknologi och se om det skulle vara lämpligt att byta KTH Formula Students nuvarande styrspindlar i aluminium (Alumec 89) till att vara gjorda av titan (Ti6AL4V) utan att öka vikten, samt inte förlora styvhet och styrka i konstruktionen.

Baserat på den nuvarande geometrin av styrspindeln för KTH Formula Students senaste bil, eV12, designades nya styrspindlar i titan med programmet SolidWorks. Denna process gjordes med hjälp av erfarenhet inom styrspindelskonstruktion och intuition, genom att analysera och förändra designen i en iterativ process.

Tre konstruktioner gjordes: en lätt version av den exisisterande, vilken var baserad på den existerande styrspindeln i aluminium, en ihålig version och en okonventionell version som utnyttjar designmöjligheter med additiv tillverkning.

För att verifiera de tre olika titankonstruktionerna utfördes det en analys av den existerande styrspindeln. Genom att använda resultatet från denna analys kunde mål för styvhet och maximal spänning sättas för den nya titanstyrspindeln.

Ingen av koncepten uppnådde de satte målen fullt ut, men värdefull insikt i design, hållfasthetslära och tillverkningsteknik erhölls. Det faktum att den specifika styvheten för titan är lägre än den för aluminium betyder att skulle vara svårt att göra en fungrande design utan användning av topologioptimeringsmjukvara, om vikt är en av de viktigaste faktorerna.

Med mer bearbetningstid skulle dessa konstruktioner troligtvis kunna möta målen.

Nyckelord: Design av Styrspindel, Additiv tillverkning, Titan, Formula Student

Examensarbete MMKB 2016:49 MKNB 088

Implementering av Additiv Tillvekning av styrspindlar för en Formula Student-bil

Sven-Ruben Böcker Kajetan Calczynski

Simon Malmström

Godkänt

2016-09-dag

Examinator

Ulf Sellgren

Handledare

Ulf Sellgren

Uppdragsgivare

Ulf Sellgren

Kontaktperson

Ulf Sellgren

(6)
(7)

Bachelor Thesis MMKB 2016:49 MKNB 088

Implementation of Additive Manufacturing in Uprights for a Formula Student Car

Sven-Ruben Böcker Kajetan Calczynski

Simon Malmström

Approved

2016-09-day

Examiner

Ulf Sellgren

Supervisor

Ulf Sellgren

Commissioner

Ulf Sellgren

Contact person

Ulf Sellgren

Abstract

This bachelor thesis focuses on the possibility to implement additive manufacturing on the upright, one of the key components in a Formula Student car. The goal was to get an insight into this manufacturing technology and to see if it would be suitable to change KTH Formula Student’s current aluminium (Alumec 89) uprights to titanium (Ti6AL4V) ones, without gaining weight and losing stiffness and strength.

Based on the current geometry of uprights for KTH Formula Student’s latest car, the eV12, new titanium uprights were designed using SolidWorks. This was done by using experience in upright design and intuition, by analysing and altering the designs in an iterative process.

Three designs were made: a lighter version of the existing one, a hollow version and an unconventional version that utilises design possibilities with additive manufacturing.

To verify the three different titanium designs, an analysis of the existing aluminium upright was performed. Using the results of this analysis, stiffness and maximum stress goals were set on the new titanium uprights.

None of the concepts fully met the set goals, but valuable insight into design, solid mechanics and manufacturing methods was gained. The fact that specific stiffness of titanium is lower than that of aluminium means that it would be hard to make a proper design without the use of topology optimisation software, if weight is one of the most important factors. With more time, the designs would likely meet the set goals.

Keywords: Upright design, Additive manufacturing, Titanium, Formula student

(8)
(9)

PREFACE

We would like to thank all people and companies that support Formula Student project all around the world, and give students the chance to explore their imaginations and improve their problem solving abilities.

Members of KTH Formula Student, especially the Vehicle Dynamics team of 2015-2016, require special thanks for their help and the ability to use their report for our work.

We would in particular like to thank Amir Rashid and the Institute of Industrial Production of KTH for their continuous support regarding manufacturing and manufacturing solutions.

Last but not least, we would like to thank our supervisor Ulf Sellgren, for his support and understanding.

Sven-Ruben Böcker Kajetan Calczynski Simon Malmström

Stockholm, September 2016

(10)
(11)

NOMENCLATURE

Notations

Symbol Description

g The gravitational constant of 9.81 (m/s2) μ Constant coefficient of friction

M Torque (Nm)

Mb Braking Torque (Nm)

F Force (N)

Fb Braking force (N)

N Normal force (N)

Fs Steering force (N)

A Acceleration (m/s2)

v Velocity (m/s)

n Rotational speed (rpm)

l Length (m)

h Height (m)

d Diameter (m)

dt Diameter of tyre (m)

r Radius (m)

rt Radius of tyre (m)

V Volume (m3)

m Mass (kg)

ρ Density (kg/m3)

σ Tensile stress (MPa)

(12)

Abbreviations

AM Additive Manufacturing

CAD Computer Aided Design

CAE Computer Aided Engineering

DMLS Direct Metal Laser Sintering

EBM Electron Beam Melting

LCA Lower Control Arm

PBF Powder Bed Fusion

SLS Selective Laser Sintering

UCA Upper Control Arm

(13)

TABLE OF CONTENTS

PREFACE ... 5

NOMENCLATURE ... 7

TABLE OF CONTENTS ... 9

1 INTRODUCTION... 11

1.1BACKROUND ... 11

1.1.1 Formula Student... 11

1.1.2 General ... 11

1.2PURPOSE ... 11

1.3DEMARCATION ... 12

1.4METHODS ... 12

2 FRAME OF REFERENCE ... 13

2.1METAL ADDITIVE MANUFACTURING ... 13

2.1.1 History of Metal Additive Manufacturing ... 13

2.1.2 Electron Beam Melting (EBM) ... 14

2.1.3 Selective Laser Sintering (SLS) ... 15

2.1.4 Practical differences between SLS and EBM... 15

2.1.5 Working with AM ... 16

2.2BASIC CAR DATA, EV12 ... 17

2.3SUSPENSION COMPONENTS ... 18

2.3.1 Control arms (A-arms) ... 19

2.3.2 Tie rod ... 19

2.3.3 Push/Pull rod ... 20

2.3.4 Upright ... 20

2.4EXISTING DESIGN USED BY KTHFORMULA STUDENT ... 21

2.5MATERIAL PROPERTIES ... 22

2.6POST-PROCESSING OF FEATURES ... 22

2.7GEOMETRIES &FORCES ... 23

2.7.1 Bump ... 23

2.7.2 Braking ... 23

2.7.3 Cornering ... 24

3 EXECUTION AND RESULTS ... 25

(14)

3.1ANALYSIS OF EXISTING DESIGN ... 25

3.1.1 Geometrical constraints ... 25

3.1.2 Mesh Parameters ... 26

3.1.3 5g Bump ... 27

3.1.4 Full braking ... 28

3.1.5 Braking while cornering ... 29

3.1.6 Conclusion ... 30

3.2DESIGN OF AMUPRIGHT ... 31

3.2.1 Key features of Upright ... 31

3.2.2 Light version of existing ... 32

3.2.3 Hollow design ... 33

3.2.4 Unconventional design... 34

3.3ANALYSIS OF AMDESIGN ... 35

3.3.1 Light Version of existing ... 35

3.3.2 Hollow Design ... 36

3.3.3 Unconventional design... 37

4 DISCUSSION OCH SUMMARY ... 38

4.1DISCUSSION... 38

4.1.1 Specific Modulus of materials ... 38

4.1.2 Post-processing after printing ... 39

4.1.3 Suspension geometry ... 39

4.1.4 Time... 39

4.2CONCLUSIONS ... 39

5 RECOMMENDATION AND FUTURE WORK ... 40

5.1RECOMMENDATIONS ... 40

5.1.1 Better technology ... 40

5.1.2 Geometry for 10-inch rims ... 40

5.2FUTURE WORK ... 40

5.2.1 Topology optimisation ... 40

5.2.2 Manufacturing of upright ... 40

6 REFRENCES ... 41

(15)

1 INTRODUCTION

This chapter describes the background, general intent, demarcation and methods for this thesis.

1.1 Backround

This chapter summarises the general background behind this thesis.

1.1.1 Formula Student

According to KTH Formula Student (2014):

Students build a single seat formula race car with which they can compete against teams from all over the world. The competition is not won solely by the team with the fastest car, but rather by the team with the best overall package of construction, performance, and financial and sales planning.

Formula Student challenges the team members to go the extra step in their education by incorporating into it intensive experience in building and manufacturing as well as considering the economic aspects of the automotive industry. Teams take on the assumption that they are a manufacturer developing a prototype to be evaluated for production. The target audience is the non-professional Weekend-Racer, for which the race car must show very good driving characteristics such as acceleration, braking and handling. It should be offered at a very reasonable cost and be reliable and dependable. Additionally, the car's market value increases through other factors such as aesthetics, comfort and the use of readily available, standard purchase components.

1.1.2 General

All authors of this report have several years of experience in KTH Formula Student and were still active members in the project during the writing of this report.

The objective is to investigate the possibility of using additive manufacturing in the production of an upright for the car that is to be used for the 2017 season, which should match the stiffness and weight of the current design.

Currently, the uprights are milled out of a single block of Alumec 89, a special aluminium alloy from Uddeholm Svenska AB which has similar properties to 7075-T6, also called aircraft-aluminium, which is commonly used in the aviation and automotive industries.

1.2 Purpose

The purpose of this project is to investigate the possibility of using additive manufacturing technology to make uprights for a Formula Student car, with the goal of achieving a decrease in weight yet having an equal or greater stiffness in comparison with the current design.

Factors such as lead time, economy and the feasibility of this manufacturing method will be taken into consideration.

(16)

1.3 Demarcation

As this is a model-based analysis, no real life tests have been made. Since this analysis is purely theoretical, some simplifications have been made.

 No topology optimisation software will be used.

 Only powder bed fusion processes are considered for manufacturing.

 All data regarding vehicle dynamics and forces will be taken from the KTH Formula Student Vehicle Dynamics group of 2015-2016.

 Fatigue will not be considered.

 Only the front uprights will be analysed.

1.4 Methods

The authors will create several concepts using the 3D-CAD software SolidWorks. These concepts will then be refined using the built-in FEA (Finite Element Analysis) tools in SolidWorks, so that desired parameters are obtained.

The current design will be analysed in different scenarios and the Von Mises stress and the displacement will then be recorded during the different scenarios. The new uprights will then be designed and analysed to match or improve upon these mechanical properties.

(17)

2 FRAME OF REFERENCE

This chapter is about the results of data acquisition and research made to be able to solve the presented task.

2.1 Metal Additive Manufacturing

This chapter introduces metal additive manufacturing and will focus on powder bed fusion, PBF, processes.

2.1.1 History of Metal Additive Manufacturing

In order to grasp the basic concepts and limitations of additive manufacturing of metals, it is beneficial to take a brief look at its history. With the massive increase of publicity, popularity and availability for consumers that plastic 3D-printing has experienced only a few years back, it is easy to regard AM technology as one in its infancy. While it is indeed true that additive manufacturing is new relative to traditional methods, e.g. subtractive manufacturing, development has undergone for several decades.

Shellabear and Nyrhilä (2004) bring up a patent application made by Pierre Ciraud in 1971 as one of the first steps towards metal additive manufacturing. Although the described method differs from the current technology, the purpose of the invention stated remains the same, albeit somewhat crude. Ciraud’s idea was to make “possible the manufacture of parts which can have extremely complex shapes, without the need for casting moulds” (Shellabear &

Nyrhilä, 2004) by the use of an energy beam and material in the form of powder.

A few years later, Ross Householder described a different way of achieving the results intended by Ciraud. A laser beam was to fuse particles together in layers, these being cross- sections of the desired part. Furthermore, Householder’s invention would also employ computers (Shellabear & Nyrhilä, 2004).

Both of the methods above required the use of laser technology, or similar, but the cost of this at the time hindered further development (Shellabear & Nyrhilä, 2004).

During the 1980s, even more patents were filed regarding additive manufacturing. Described were methods utilizing previous concepts, such as Householder’s, with lasers focusing energy on layers of either liquid or powder material. These technologies, now receiving names like Selective Laser Sintering, were developed and tested by more or less independent actors.

Despite having success with polymers, though, metals proved more difficult to work with (Shellabear & Nyrhilä, 2004).

The 1990s saw the commercialisation of metal additive manufacturing after fruitful research on metal powders. Direct Metal Laser Sintering, DMLS, was the first functioning method of 3D printing metal, with an initial layer thickness of 100µm. By the year 2001, the thickness had been reduced to as little as 20µm. Machines using similar techniques of fusing metal powder were also created during the first years of the new millennium, such as Selective Laser Melting, SLM, and Electron Beam Melting, EBM (Shellabear & Nyrhilä, 2004).

One of the primary reasons that additive manufacturing has become so popular is the reduction in time for product development. CAD models can be relatively easy to build in a 3D printer thanks to the ability to print different parts in the same build, or including

(18)

connecting joints in an assembly. While it may take some time to arrange the parts in an efficient way, the total time can be remarkably less than before. CNC machining, on the other hand, requires intricate preparation like programming and calibration for the various stages, something AM can ignore to a large degree (Gibson, et al ., 2010).

In their paper written in 2004, over ten years ago, Shellabear and Nyrhilä state that the market potential is unknown, given how recently the various machines had been released. There were also legal insecurities related to the rights of the patented methods (Shellabear &

Nyrhilä). Moving forward one decade, it is safe to assume that metal additive manufacturing has indeed had an impact. The market for 3D printing, of all materials, was estimated to be over 5 billion USD in 2015, after a growth of 1 billion USD from 2014, and it has maintained a steady increase of around 30% each year (3ders.org, 2016).While polymer-based machines and materials account for much of this, metal printers have seen a greater relative increase in sales in recent years, with nearly 800 units sold in 2014 alone according to Dunham (2015).

This is accredited to the machines moving from merely institutional research, prototyping and small scale production towards being a profitable part of manufacturing. While there are a number of companies providing metal printers, it is commented by Dunham that they are not necessarily direct competitors due to the complexity of the various technologies and their appropriate applications (2015).

2.1.2 Electron Beam Melting (EBM)

Electron Beam Melting uses, as the name suggests, an electron beam as the energy source in a powder bed fusion process. The technology, developed at Chalmers University of Technology in Sweden, resembles that of cathode ray tubes which can be found in older television sets. In the Arcam A2X (a AM machine available at KTH), the electrons are emitted by a tungsten filament heated by an electric current to a temperature high enough to let the electrons escape. These are accelerated in an electric field and the beam can be controlled by a focusing coil, where spot size is set, and a deflection coil, directing the beam’s x-y motion. When the electrons hit the powder bed, and the metal particles, their kinetic energy is absorbed as heat, fusing the particles together. Even though the electron has a mass of just 9.1 × 10-31 kg, it can reach a velocity up to 0.4 times that of light in the machine (Arcam, 2011).

The high energy output at low costs of EBM technology has so far been superior to the levels reached by using a laser, but the latter does not induce the same type of issues with the particles. As the electrons, themselves being particles with mass, have a negative electric charge, this is transferred to the metal particles upon contact. Two negatively charged particles will exert a repulsive force on each other, and this cannot surpass the force of friction holding each particle in place. Also, the incoming electrons can be deflected for the same reason, further adding to the problem. This is avoided by having a less focused beam and scanning the surface in specific ways (Gibson, et al., 2010; Arcam, 2011).

Because the powder particles become negatively charged, the part being manufactured must be conductive and attached to the base of the build platform by support structures (Gibson, et al ., 2010).

Inside the build chamber is a vacuum, as an inert gas would contain atoms the electrons could

(19)

2.1.3 Selective Laser Sintering (SLS)

Selective Laser Sintering, SLS, is a term used for powder bed methods using lasers as the thermal source. Printable material is polymers, metals and ceramics, due to the fact that the energy beam consists of photons rather than electrons and the build particles receive no electric charge. This also means that energy can be concentrated to a smaller point, as opposed to EBM, giving the final part a better surface finish, and that an inert gas can be used inside the build chamber to prevent oxidation. The laser is directed using lenses and mirrors.

A term typically used for SLS when metal powder is used is Direct Metal Laser Sintering, DMLS. Originally developed by EOS, Electro Optical Systems, variations of DMLS have been created and made available to the market (Gibson, et al., 2010).

The word sintering implies that the particles are fused together without fully melting. It is not uncommon, though, to refer to the process as melting, which is not entirely incorrect since full melting can be done with lasers too. There are different processes, but the most prominent one is liquid phase sintering (Gibson, et al., 2010), used in DMLS machines (Binelli, et al., 2011), where some parts of the material are melted but others remain solid, the former acting as a binder. The solid structural material can be completely separate from the binder or these can be combined in one single powder particle. Structural particles may also be coated with the binder material.

2.1.4 Practical differences between SLS and EBM

It is deemed appropriate by the writers that both laser and electron technologies are mentioned, in the case of an opportunity to use a machine other than the Arcam A2X EBM printer. Although they follow the same basic principles, EBM and DMLS differ on certain practical key points that are relevant for members of KTH Formula Student.

Many alloys are printable. EOS, developer and leading manufacturer of laser sintering machines, have parameter settings for desired material properties for metals other than titanium, such as aluminium and steel (EOS, 2015). Arcam, though, only have a few standardised materials where appropriate settings are available. Of course, it is possible to attempt using other metals (Arcam can provide support, too) (Arcam, 2014), but this would be more suitable as a research project. Since the uprights are meant to be manufactured at the university, this limits the material choice.

As previously mentioned, particles are fused in liquid phase sintering in the DMLS process.

With EBM the powder particles are fully melted, and previously solidified material is melted as well, making it possible to create more dense structures. Whereas heat is more concentrated to the actual build in a DMLS chamber, a more even temperature is kept in the EBM powder bed. The result of the above is that DMLS produces a clearly layered microstructure and EBM one resembling cast metal. It also leads to less residual stress in the EBM machine meaning less material needed for support structures to counteract this (Gibson, et al ., 2010).

Because the electron beam is less focused than its laser counterpart, the resulting finish, layer thickness and resolution are all affected. These tend to be larger and coarser for EBM than any of the laser processes. In general, EBM is a faster process than DMLS under similar conditions, such as layer thickness of the build. Part of this is because the electron beam can be controlled using electromagnetic fields alone and several melt pools can be maintained at the same time. Manufacturing time may be an issue depending on the availability of a

(20)

machine. For example, a partner may be unwilling to provide support if the company’s lost production time is too costly.

2.1.5 Working with AM

Additive manufacturing is a very broad term that includes technologies differing greatly in some ways. According to Gibson et al., though, there is a general approach to be taken when working with AM. The first step is to create a 3D model in a CAD system, which is the subject of this report. Next, the CAD file is converted to an STL file, a format used by AM machines, and support structures are added. The file is then transferred to the machine and parameter settings like scanning speed, layer thickness and build temperature are entered.

Before the part is built, a simulation of the process is done to ensure no problems will arise.

When this is finished, it must be detached from the build platform and excess material, metal powder in our case, removed. Post-processing is often needed before the part is finished and ready for use. One of the advantages of 3D printing in the last stage, application, is that several parts can be made in the same build with joining links, reducing assembly time, but it is unlikely that KTH Formula Student will do this.

There are three programs specifically used for Arcam machines. Magics converts CAD files to STL format, arranges parts and adds supports. EBM Build Assembler creates Arcam Build files using the information imported from STL files. These are then exported to EBM

Control, in which the process is simulated and settings applied. It also manages the actual build process (Arcam, 2011).

How a part is to be oriented in the build chamber should be considered in the design phase, as this can have great effect on production time and the resulting properties. The duration of the manufacturing process depends on the height of the total build volume; parts printed lying down will require fewer layers and thus be finished sooner. It is inadvisable to do this for certain geometries, though. Since the parts are built in relatively thick layers, stair-step patterns can be seen on round or curved sides built perpendicular to the layer plane. This is not a desirable quality for many components, and these function better when printed upright.

It should be noted, too, that there are often complex shapes in more than one direction to be accounted for (Gibson, et al., 2010).

The space available for a build in a machine is also a limiting factor. For the Arcam A2X, the maximum build area is 200 × 200 mm, and height is limited to 380 mm (Arcam, 2014). A part that would be best printed face-down may have to be printed standing up instead due to it exceeding the dimensions of the build area, resulting in poor properties of the final product.

Creating hollow structures is possible with AM, and doing this can reduce both production time and weight of the final component. Hollow parts will still contain metal powder that has to be removed, meaning that holes or a similar way of getting the powder out must be

designed as part of the component.

(21)

2.2 Basic Car Data, eV12

Table 1 holds the dimensions and performance figures related to KTH Formula Students latest car, the eV12, shown in figure 1.

Figure 1. A rendering of eV12 Table 1. Basic car data

Dimensions

Data Value Comment

Length 3200 mm

Width 1300 mm Length between outer face of tyres.

Track Width 1100/1080 mm Front/Rear. Length between mid- plane of tyres

Height 1085 mm

Wheelbase 1530 mm Distance between Front and Rear

axel

Weight 300 kg 220 kg vehicle weight, 70 kg driver

weight, 10 kg equipment weight Weight Distribution,

Stationary

45/55 Front/Rear

Weight Distribution, Braking

80/20 Front/Rear

Tyre Diameter 520.7 mm (20.5 in) Hoosier FSAE Racing Tire, 43163 (Hoosier, n.d)

Tyre Width 177.8 mm (7 in) Hoosier FSAE Racing Tire, 43163 (Hoosier, n.d)

Performance Acceleration (0-100km/h) 3 seconds

Top Speed 180 km/h

Max power 80kW (107 hp)

Max torque 1150 Nm

Max voltage 567 V

(22)

2.3 Suspension Components

The suspension design of eV12 features a short long arms (SLA) geometry. This refers to the upper and lower a-arms being of different lengths in order to obtain specific kinematic properties. Figure 2 illustrates some of the suspension components that are mentioned in this report.

(23)

2.3.1 Control arms (A-arms)

This component, shown in figure 3, is one of the links between the chassis and upright. It has two nodes in the chassis and one in the upright, giving it the shape of the letter A, hence the name. The A-arms are not adjustable and transfer the majority of all forces generated by the tyre to the chassis.

The arms are made out of tubes welded to turned steel cylinders (the nodes). In each of the nodes sits a spherical bearing, letting the control arm move in the vertical direction. There are two general categories for control arms: Upper Control Arms, UCA for short, and Lower Control Arms, LCA.

Figure 3. UCA (note that LCA does not have a mounting bracket).

2.3.2 Tie rod

The tie rod, seen in figure 4, connects the steering system to the upright and transfers all steering forces between the steering wheel and the tyre. They are adjustable and provide the ability to change the toe angle. The tie rod is made out of a steel tube with welded ends. One side is threaded to allow attachment of a rod end, while the other side allows for attachment of an aluminium end for shimming.

Figure 4. Tie rod.

(24)

2.3.3 Push/Pull rod

The push and pull rods link the uprights to the vehicle’s dampers and springs through rockers.

It transfers the normal load from the tyre and displaces the wheel vertically according to the motion-ratio of the damper. The push/pull refers to the direction of actuation. For the front suspension of eV12, pull rods are used. The pull rod is adjustable to allow the setting of different ride heights. The pull rod has the same design as the tie rod.

2.3.4 Upright

A key component in a suspension system is the upright, seen in figure 5. It is a non-rotating part that links the other components together. There are four parts connected to the upright:

LCA, UCA, the tie rod and the wheel hub, the latter being housed by the upright via deep groove ball bearings allowing the wheels to rotate. These connections are required to achieve the desired kinematic properties and degrees of freedom. Forces at the tyre’s contact patch are transferred trough the wheel, its hub and the wheel bearings before reaching the upright.

A description of the way the upright is manufactured can be found in 2.4 Existing Design used by KTH Formula Student.

Figure 5. Upright assembly.

(25)

2.4 Existing Design used by KTH Formula Student

The current design, seen in figure 6, is milled out of a single block of Alumec 89, provided by Uddeholm Svenska AB, a high-strength aircraft grade Aluminium with similar material properties to 7075-T6. The weight of the current design is 600 grams for the front uprights, according to CAD data.

Front View Rear View Side View

Figure 6. Different views of existing upright.

The lead time of all four uprights is around 2-3 weeks when manufactured in the workshop at KTH Institute of Industrial Production, IIP. The rear and front uprights differ in design, and even though the left and right ones are mirrored versions of each other they require their own CAM (Computer Aided Manufacturing) codes, meaning that the entire manufacturing process is very time-consuming.

The uprights need to be milled from two sides. First, they are milled from the rear view in figure 7, then turned over and put into a jig that centres them on the bearing seat, and later have their front side milled. The last process is to drill all holes on the sides, done in a manual mill.

Rear View Illustration of jig

Figure 7. How the existing uprights are milled.

(26)

2.5 Material Properties

Table 2. Material Properties. (Arcam, n.d; Uddeholm, 2011)

Mechanical Properties Arcam Ti6AL4V Uddeholm Alumec 89

Yeild Strenght (Rp0,2) 950 MPa 520 MPa

Tensile (Rm) 1020 MPa 560 MPa

Elongation 14% 9,6%

Hardness 310 HB 140-180 HB

Modulus of Elasticity 120 GPa 71.5 GPa

Densisty 4.506 kg/m3 2.83 kg/m3

In order to set a benchmark for material properties, a safety factor of 1.33 was chosen when designing for stress. This means that the maximum stress level does not exceed 75% of the yield strength, 0.75𝜎𝑠 (Gedeon, 2005). With the help of table 2, the maximum stress allowed would be around 710 MPa for a titanium upright, and 390 MPa for one made in aluminium.

2.6 Post-processing of Features

Even though it is possible to create highly complex geometries with additive manufacturing, some features may need to be post-processed. This post-processing can be done by milling, turning, hardening, or similar. For the component in question, post-process milling is required due to the rough surface finish of the 3D-printed part. For the Arcam A2X, surface roughness for vertical and horizontal surfaces varies from 𝑅𝑎 25 to 35.Features such as the bearing seats and mounts are affected by this. SKF recommends that the surface roughness should be in the range of 𝑅𝑎 0.8-1.6 for their bearings (SKF n.d). To meet these requirements, the parts have to be milled or turned. Milling results in a surface roughness of 𝑅𝑎0.8, and turning a roughness as low as 𝑅𝑎0.4 (surface roughness table can be found in Appendix A).

While turning the bearing seats would lead to the lowest surface roughness, there is currently no machine at KTH IIP with the required abilities, and the unfinished upright would have to be sent somewhere else (further increasing lead time). Instead, the better alternative is to use the Hermle C50, a 5-axis CNC mill at KTH IIP. It can measure the part and do all the required milling features in one setup. This would save both time and money. One problem with this method is that a milling machine uses interpolation to create a circle, meaning that the feature may end up not being perfectly round. However, the bearing seats for the last two cars’ uprights have been milled without any problems occurring.

(27)

2.7 Geometries & Forces

An upright will be exposed to many different kinds of loads during a driving session. To simplify, the possible scenarios can be summarized into three categories: bump, braking, and braking while cornering. A illustration of the forces acting on the system are shown in figure 8.

Figure 8. Free body diagram of the front wheel assembly.

2.7.1 Bump

In the case of a bump, the normal force experienced by the car is increased. One usually refers to the magnitude of bumps in g-forces, where 1 g is the normal force acting upon the object when stationary. In this case, 1g is about 675 N for each of the front uprights, assuming a total weight of about 3000 N and weight distribution of 45/55, front to rear.

2.7.2 Braking

Braking implies that the mechanical brakes of the car are fully engaged. A closer inspection reveals that there are two major loads acting upon the upright during this scenario. One is the braking torque generated by the brake calipers that act upon where the brake caliper is attached to the upright. Studies by KTH Formula Student have shown that this breaking torque can have a magnitude of up to 450 Nm (Sanchez, et al. 2016, 92-96).

The second force, acting upon the bearing seat, is the result of the braking torque. This force is located at the contact patch of the tyre and is calculated by dividing the braking torque by the tyre radius (1).

𝐹𝑏 =𝑀𝑟𝑏

𝑡 (1)

(28)

2.7.3 Cornering

According to the paper written by the 2015 Vehicle Dynamics group of KTH Formula Student, the total force acting upon the rack and pinion joint of the steering is 1118 N (Sanchez, et al. 2016, 91). Using this data, the steering torque that will act upon the upright can be calculated.

The kingpin axis, which is the axis that goes through the attachment points of the wishbones, is shown in figure 9. The distance between the axis and the point of the steering force is 60mm. This means that:

𝑀𝑠 = 𝐹𝑠 ∗ 𝑟𝑠 = 1118 ∗ 60 ∗ 10−3= 67.8 𝑁𝑚 (2) The scenario stated above does not entirely match the one in reality. When the upright rotates around its kingpin, the moment arm (distance shown as a black line in Figure 9) will decrease. This is due to the kinematic properties of the steering geometry, but not something that will be accounted for in this analysis. To ascertain that the upright will still perform during the worst-case scenario, this torque was set to be 70 Nm (2) during cornering, which is the highest torque during the steering scenario.

(29)

3 EXECUTION AND RESULTS

This chapter covers the way the task was executed and the results that followed.

3.1 Analysis of Existing Design

The following chapter is an analysis of the existing upright made by KTH Formula Student.

This analysis is to set a benchmark for the new upright that will be manufactured with AM.

Three different scenarios will be looked at: 5g bump, full braking and braking while cornering. All of the forces have a safety factor of about 1.5 compared to what they would be in reality.

3.1.1 Geometrical constraints

To be able to have an initial starting point for the job at hand, an analysis of the existing design was performed. As mentioned above, all work was made using SolidWorks.

To start off, the full assembly of the upright was acquired from KTH Formula Student, and only the parts up to the attachment of the steering rod, lower and upper control arm (LCA and UCA) were left in the assembly. This was not only to make the calculations more reliable, but also to make them easier. The assembly consisted of 8 parts in total.

In figure 10 one can see the constraints, connections and fixtures that where applied to the model. Component contact was set to “No Penetration” with a friction of 0.2. This is due to the fact that most parts where made out of Aluminium.

The green arrows indicate constraints.

Constraint 1, also called UCA, and was set to not allow translation in any direction.

Constraint 2, also called LCA was set to not allow translation in X- and Y-Direction

Constraint 3, which is the steering attachment point, was set not to allow translation in Y- direction.

The purple arrows indicate forces and torques.

Force 4 and 5 (purple) are Bearing Loads that are transmitted from the tyre, through the hub to the upright. Force 4 is in the X-direction and pointing opposite of the driving direction, in other words backwards. Force 5 is in the Z- direction and pointing upwards. Torque 6 is the

Figure 10. Constraints and forces acting on the upright.

(30)

torque acting upon the caliper mount from the brake disk. Torque 7 (light blue) is the torque that is generated by cornering. This torque is revolving around an axis that is parallel to the ground plane and is acting upon each of the bearing seats.

There is also another torque acting upon the upright while cornering and braking, which twists the upright around its kingpin axis. For simplistic reasons, the torque is displayed in figure 14.

The blue, unnumbered, connections are bolts in size M4, M5 and M6.

A series of analyses were preformed, were the magnitude of the forces was changed to mimic different driving scenarios.

3.1.2 Mesh Parameters

Figure 11 shows the mesh parameters that are set in SolidWorks.

Figure 11. Mesh parameters in SolidWorks.

(31)

3.1.3 5g Bump

This scenario is quite simple. The car hits a bump that increases the normal force by 5 times.

This is an extremely rare thing to occur, but nevertheless it is a necessary parameter that the upright should withstand (Flickinger, 2014).

The force was set to be upwards (force 5, in figure 10) and had a magnitude of 3750 N.

The results can be observed in figure 12.

Figure 12. 5g Bump results on existing upright.

(32)

3.1.4 Full braking

The scenario implies that the car was going straight and the brakes were fully engaged.

The force acting upwards (z-direction) had a magnitude of 1200 N; the brake torque was 450 Nm which resulted in a bearing load of 1800 N backwards (x-direction).

The results can be observed in figure 13.

Figure 13. Full braking results on existing upright.

(33)

3.1.5 Braking while cornering

As mentioned above in chapter 3.1.1 Geometrical constraints, a steering torque around the Kingpin axis was added to simulate the aligning torque of the wheels. Since figure 10 would be too cluttered with this torque being shown, it is shown here in figure 14.

Figure 14. Wheel aligning or steering torque acting upon upright.

The scenario implies that the car was cornering and the brakes were engaged.

The force acting upwards (z-direction) had a magnitude of 1000 N, the Brake torque was 300 Nm which resulted in a bearing load of 1200 N backwards (x-direction). In addition, a steering torque was also acting around the kingpin axis of 70 Nm and a cornering torque of 220 Nm. The results can be observed in figure 15.

Figure 15. Braking while cornering results on existing upright.

(34)

3.1.6 Conclusion

All scenarios result in a stress that is under the design stress of 0.75𝜎𝑠 presented in chapter 2.5 Material Properties. It is not surprising that the highest stress is present while cornering and braking at the same time, and the highest displacement during full braking. This design will not be further analysed.

(35)

3.2 Design of AM Upright

As previously stated in this report, the models that will be designed for AM are based on general CAD and solid-mechanics knowledge and experience. The designs are made in an iterative manner, meaning that the models undergo finite element analysis between design iterations followed by geometrical corrections. In this report, only the last iteration of designs is shown. The goal is to achieve approximately the same mechanical properties as the existing design (see chapter 3.1 Analysis of Existing Design). Three different parameters will be looked at: Von Mises stress, displacement and weight.

The aim is a maximum of 0.75𝜎𝑠 for stress, 0.4 mm for displacement, and a weight of 600 grams or less.

To meet these requirements, three different concepts were created with some of the benefits posed by AM described in chapter 2 Frame of Reference. In one design, material was removed from the original upright model in ways not possible with subtractive manufacturing. Another design was based on the possibility of creating hollow structures.

The last approach differed from the first two, as these were using the aluminium upright as a reference baseline. It was an attempt at creating an entirely new design without regard to the limitations posed by subtractive manufacturing, but rather with AM in mind.

3.2.1 Key features of Upright

In order to be able to make the design process as smooth as possible, a baseline or base-CAD file was created with the key features of the upright. In figure 16, the baseline designs of the so called Hollow (chapter 3.2.3) and Unconventional (chapter 3.2.4) designs are presented.

The Light version of existing (chapter 2.3.2) did not require a baseline since it is derived from the original aluminium design.

Four essential features were identified: the bearing seat (the large cylinder in the middle in figure 16), the LCA attachment point (lowest geometry), the UCA attachment point (top bracket) and the brake caliper attachment bracket (two small squares with holes).

Figure 16. To the left, key features of 3.2.3 Hollow design. To the right, key features of 3.2.4 Unconventional design.

(36)

3.2.2 Light version of existing

This is a very simple design compared to the original one. The material of the existing aluminium upright was changed to titanium and material was then removed so that the model’s weight was as close to that of the original upright as possible. The weight of the titanium upright was about 578 grams.

This was made possible by the ability of AM to omit unnecessary material from areas where it would not be possible with subtractive manufacturing. As seen in figure 17, a clear example of this is the beams stretching from the bearing seat to the top attachment point of the upright.

Front View Rear View Side View

Figure 17. Different views of Light Version in titanium of existing upright.

(37)

3.2.3 Hollow design

The general idea behind this design, seen in figure 18, is that a solid geometry that has exceptionally good torsional stiffness in relation to its weight is a pipe or a hollow beam.

Therefore, this concept tries to utilise the parameters of a tube. One problem with this design is that the support material has to be removable from the inside of the upright, meaning that open channels have to be integrated into the design on the side, and between the bearing seats. Another challenge is the way the tie-rod attachment point is constrained. The design of this bracket is based on an I-beam, where the top bracket provides good support for loads in all directions. The bottom bracket gives low support for vertical loads in order to avoid stress disturbances in the LCA joint. This version weighs 565 grams.

Front View Section View Side Section View

Figure 18. Different views of Hollow design.

(38)

3.2.4 Unconventional design

This design, shown in figure 19, uses the idea of only having material in the necessary places, therefore saving weight and improving stiffness. A different baseline was used on this design to give more freedom of how to design brackets and where to add material (see chapter 3.2.1 Key features of Upright).

Like the previous concept, this is meant to feature hollow structures and beams, and therefore suffers from the same problem: removal of excess powder. Holes need to be drilled in each of the hollow sections, and this would alter the mechanical properties of the structure.

This design was set aside due to the time-consuming work process and to the fact that it would be very difficult to achieve the goals set for mechanical properties without any topology optimization software. The weight of the final iteration is 626 grams, without hollow structures, according to CAD data.

Front View Rear View Side View

Figure 19. Different views of Unconventional design.

(39)

3.3 Analysis of AM Design

All three concepts were analysed with the same load scenarios as those of the aluminium upright. Shown below are the worst-case scenarios, i.e. the ones where the biggest stress and largest displacement are present. The other results are presented in Appendix B.

3.3.1 Light Version of existing

The results show that this design meets the set requirements for von Mises stress, presented in section 2.5 Material Properties. However, the displacement during full braking and braking while cornering (shown in Appendix B) exceeds the goal set in section 3.2 Design of AM Upright.

During braking while cornering, it is only the brake mount that is over the set displacement limit. This means that it could be redesigned to fit the parameters without addition of too much material. That, however, is not the case for the full braking scenario, seen in figure 20, due to the fact that so much of the geometry exceeds the set goal. This means that a more thought-out redesign of the upright would be required, and with only 22 grams to spare it would be a difficult task to do.

5g Bump Full braking

Figure 20. To the left, Von Mises stress during 5g Bump. To the right, displacement during Full braking on Light version.

(40)

3.3.2 Hollow Design

The hollow design also meets the stress requirements in all load scenarios. As can be seen in figure 21, though, the displacement was well over the limit in the case of Braking while cornering. This upright only met the displacement criterion for 5g bump (see Appendix B).

Full braking Braking while cornering

Figure 21. To the left, von Mises stress during Full braking. To the right, displacement during Braking while cornering on the Hollow version.

(41)

3.3.3 Unconventional design

Only one load case, braking while cornering, was analysed when designing this upright.

Running just one analysis on a high-performance workstation took almost 20 hours, likely due to the complex geometries of the part. Because the work method was to analyse and alter the models in an iterative way, an unconventional design meeting the requirements was never finished.

The results presented in figure 22 show that the upright, like the previous ones, could manage the load scenario overall except for in one area when looking at stress. Displacement is a bigger problem with this design, and unlike the others, adding material or creating beam structures at exposed points is not easily done without completely altering the geometry.

Figure 22. To the right von Mises stress and to the left displacement, during braking while cornering.

(42)

4 DISCUSSION OCH SUMMARY

In this chapter the results are summarized and discussed, so that conclusions could be made.

4.1 Discussion

In this chapter the results are discussed.

4.1.1 Specific Modulus of materials

A common trend can be seen with all results: the printed titanium components do not have similar stiffness as the milled aluminium ones. The structural geometry might be a cause, but something that has been noted is the specific modulus, also known as specific stiffness, of titanium. The specific modulus of a material is the elastic modulus, 𝐸, per mass density. Put crudely, it is how stiff a material is in relation to its weight. Figure 23 presents the specific modulus/stiffness of common materials.

Figure 23, Specific stiffness of common materials. (Tetrafix n.d)

The figure states something that is well known, namely that steel, aluminium and magnesium have the same specific stiffness. This is why many components, like car rims, that were previously made out of steel can be made out of aluminium or magnesium today.

Since titanium has a lower specific stiffness than the materials listed above, it cannot have the same weight and stiffness as an aluminium upright with a similar design. The only way to lower weight is to apply complex geometries, something that can be done with additive manufacturing.

Removing material on a component made out of titanium will also have a greater impact on its stiffness than on an aluminium one.

(43)

4.1.2 Post-processing after printing

Even though time would be saved during the manufacturing of the upright by printing it, compared to milling, a lot of other factors would have to be looked at. With today’s technology, the bearing seat would still have to be processed.

To be able to post-process the printed geometry, it would have to feature attachment points for a jig. This is so that it could be processed in the mill, as well as being able to orient the part in the machine correctly.

4.1.3 Suspension geometry

Due to the facts stated above about specific stiffness another discussion point can be made, that the current suspension geometry is not suitable for printing. If the lengths between attachment points would be shorter, lower forces would be acting upon the upright. This, however, would not be possible for 13-inch rims due to kinematic properties of the suspension according to the team responsible of the vehicle dynamics of the car. The only way to do so is to change to 10-inch rims. This would require and allow the upright to become smaller, which in the end reduce the lengths between attachment points.

From what has been seen in Formula Student, titanium uprights for 13-inch rims has rarely been used, but it is more frequent for 10-inch rims.

4.1.4 Time

Time is constantly an obstructive factor for KTH Formula Student in many ways. A good approach would be to use topology optimisation software, but this could in turn pose a problem time-wise. The team’s high-performance workstations have limited simulation-time due to the fact that other components on the car need to be analysed as well.

4.2 Conclusions

There are several factors that hinder the use of AM for a suspension setup that features 13- inch rims. First of all, the design process is much more complicated than that of the aluminium uprights, which is a well-proven method. This results in a long and tedious design process, which would have to be done with topology optimisation software.

If a design can be created that meets the goals set in chapter 3.2 Design of AM Upright, the chance of it being lighter than 600 grams is very slim considering the specific stiffness of titanium, discussed in chapter 4.1.1 Specific Modulus of materials.

This leads to the manufacturing planning and execution. Problems such as building orientation, warping and other situations related to building geometry have not been looked at. As of now, a clear path on how to process the key features of the design after printing is not known. What is certain is that a jig of some kind would be necessary to be able to hold the piece while processing, as well as probing points for part orientation.

To add to these points, economy has not been discussed thoroughly. A quick look into it shows that titanium powder is much more expensive than a block of aluminium.

As a conclusion, implementing additive manufacturing for uprights for a Formula Student car requires more research into use of topology optimisation software, a clearer manufacturing plan, money and probably a new suspension geometry.

(44)

5 RECOMMENDATION AND FUTURE WORK

In this chapter recommendations and future work are presented

5.1 Recommendations

In this chapter recommendation based on the work done are presented.

5.1.1 Better technology

The surface finish poses a problem for AM. A good surface finish is often essential to the function and performance of a part. One example is the bearing seat of the upright, where it is very important to have good tolerances and perfect concentricity. Therefore, better technology that would give better tolerances and a better surface finish would benefit this construction greatly.

5.1.2 Geometry for 10-inch rims

As previously stated in chapter 4.1.3 Suspension Geometry, it is recommended to change to another solution of wheels and tyres, consisting of 10-inch rims, if implementation of AM- uprights is desired.

5.2 Future Work

In this chapter, advice is given about future work.

5.2.1 Topology optimisation

A better study of the forces, as well as using topology optimisation programs, would result in a better design than one based on intuition. The use of this type of software could greatly improve the strength and stiffness of the part, without compromising the weight.

5.2.2 Manufacturing of upright

A study that focuses on the way parts are manufactured in a specific machine, for example Arcam A2X, to lay a baseline of how to or how not to position a part during building based on building time, finish and strength of finished part.

A study that focuses on finding a good way to post-process key features on parts that have been produced with AM. Topics such as jigging, finding reference points, cutting speeds, cutting tools and alike would be researched and tested.

(45)

6 REFRENCES

3ders.org, “Wohlers Report 2016 reveals $1 billion growth in 3D printing industry”,

http://www.3ders.org/articles/20160405-wohlers-report-2016-reveals-1-billion-growth-in-3d- printing-industry.html, accessed 2016-09-11, 2016.

Arcam,”Arcam EBM User’s Manual”, Rev.1, 2011 Arcam,”JUST ADD”,

http://www.arcam.com/wp-content/uploads/justaddbrochure-web.pdf, accessed 2016-09-11, 2014.

Arcam,”Ti6Al4V Titanium Alloy”,

http://www.arcam.com/wp-content/uploads/Arcam-Ti6Al4V-Titanium-Alloy.pdf, accessed 2016-09-11, n.d.

Binelli A., Perez A., Jardini A, Filho A, “Direct Metal Laser Sintering (DMLS): Technology for design and construction of microreactors”, 6th Congresso Brasileiro de engenharia de fabricação, 2011

Dunham S., “Quantifying an Explosion: Today’s Metal Additive Manufacturing Market”, http://www.disruptivemagazine.com/features/quantifying-explosion-today%E2%80%99s- metal-additive-manufacturing-market, accessed 2016-09-11, 2015.

EOS, “EOS Materials”,

http://www.eos.info/material-m, accessed 2016-09-11, 2015.

Flickinger E.D, “Design and Analysis of Formula SAE Car Suspension Memebers”, 2014 Gedeon M., “Yield Strength and Other Near-Elastic Properties”,

http://materion.com/~/media/Files/PDFs/Alloy/Newsletters/Technical%20Tidbits/Issue%20N o%2047%20-%20Yield%20Strength%20and%20Other%20Near-Elastic%20Properties.pdf, accessed 2016-04-28, 2005.

Gibson I., Rosen D., Stucker B.,”3D Printing, Rapid Prototyping, and Direct Digital Manufacturing. “Additive Manufacturing Technonlogies, Second Edition, 2010 Hoosier, “FSAE Racing Tire“,

https://www.hoosiertire.com/Fsaeinfo.htm, accessed 2016-05-07, n.d.

KTH Formula Student,“Formula Student “,

http://www.kthformulastudent.se/index.php/formula-student, accessed 2016-08-01, 2014.

Sanchez G., Orzi S., Sanz S., Dabhi M., Gabriele S., Massolo M., Shekhar R., “SD 2229- 2230 - Vehicle Dynamics Project Course”, Master's Programme: Vehicle Engineering KTH, 2016.

(46)

Shellabear M., Nyrhilä, O.,”DMLS – Development History and State of the Art”, LANE 2004 conference, Erlagen, Germany, 2004

SKF, “Surface finish for bearings”, http://www.skf.com/group/products/bearings-units- housings/roller-bearings/principles/design-considerations/radial-location-bearings/surface- roughness-of-seats/index.html, accessed 2016-04-15, n.d.

Tetrafix, “SUPERIOR PERFORMANCE BY USING CARBON FIBRE”, http://www.tetrafix.se/en/carbon-fibre.html, accessed 2016-07-28, n.d.

Uddeholm, “Product brochure Uddeholm Alumec 89”,

http://www.uddeholm.com/files/PB_alumec_english.pdf, accessed 2016-03-01, 2011.

(47)

APPENDIX A: SURFACE ROUGHNESS TABLES

(48)

APPENDIX B: RESULTS OF OTHER SCENARIOS Light version

5g Bump

Cornering while braking

(49)

Hollow Version

5g Bump Full braking

5g Bump Braking while cornering

References

Related documents

In this concept shown in Figure 2.5 each of the tie rods on the car are replaced entirely with electrically controlled linear actuators.The actuators would be controlled

Keywords: Additive manufacturing (AM), electron beam melting (EBM), total hip replacement (THR), orthopedic implants, digital design, computer aided design (CAD),

It is due to these diculties that methods of simulation such as computational uid dynamics (CFD) have been developed to aid the research and it is the objective of this thesis

Figure 30 shows the total cost of AM when accounting for added value for all design cases with the base values.. The added values have little impact on the total cost because the

The design of the rear wing was based on the current rear wing on the KTH Formula Student vehicle.. The current rear wing was slightly redesigned for this project to

While there was no formal requirement on qualification for the turbine (being a demonstrator), both external and internal reviews defined verification requirements according

LINA BERGLUND, FILIP IVARSSON, MARCUS ROSTMARK. KTH ROYAL INSTITUTE

The electron beam melting additive manufacturing method is promising for the repair of damaged Inconel turbine blades since the technology allows to produce free